8VSB has to go..........................
DTV News:- Department of Personal Enlightenment - I: I participated in the
Sinclair Broadcasting DTV modulation tests in Baltimore on Friday. I
say "participated in" rather than "observed" because the Sinclair
personnel were totally open to my doing whatever I wanted and to giving
me hands-on control (little did they suspect...).
These were (and still are) the parameters: They are operating
under special temporary authorization on channel 40 at about 50 kW at
about 1250 feet. The COFDM and 8-VSB alternate on the same transmitter
(a call is made to the transmitter to have the modulation switched),
same feed line, same transmitting antenna (to the south only), same
receiving antennas, same receiver RF amplifier. Any channel anomalies
affect both forms of modulation identically. The high-power amplifier
is one that Sinclair normally uses for its channel 45 (WBFF) analog
station in Baltimore, so the analog station is running at half power
during the tests. They are meeting the ATSC bandpass criteria but not
quite the FCC's, but that's only because they don't want to deal with it
just for the testing.
The receiving setup consists of a cheapo UHF dipole bow-tie (as is
included with most TVs), a Radio Shack dual bow-tie with reflector (a
$15.99 special) with excessively long twin-lead (see later), some RF and
video switching gear, two 8-VSB receiver/decoders (Panasonic and
Pioneer), two COFDM receiver/decoders (Nokia and NDS, the latter an
older, professional unit), a Sony picture monitor, and a spectrum
analyzer. The antennas are mounted on a tripod, the ultra-cheapo dipole
at about seven feet up, the directional at about five feet up. There is
an azimuth scale for noting antenna rotation.
The COFDM parameters were (and still are) 64-QAM, 3/4 FEC, 1/8
guard interval, "2k" (1705) carriers, and roughly 18.7 Mbps payload.
Both modulators were fed streams from servers. My personal observations:
- There are significant differences between the receiver/decoders
within each form of modulation. I wish they had had more
receiver/decoders. Sometimes the differences were related to
sensitivity, sometimes to ability to deal with other channel defects. I
did not judge picture quality; that wasn't the purpose of this test. I
will not point out which specific receivers did what, because I think it
is unfair to judge a sample of one. For what it's worth, three of the
receivers had near-identical sensitivity; the fourth was an 8-VSB
receiver not as sensitive as the others.
- A signal-strength meter or indicator on a receiver/decoder will
probably be all but useless as a means of indicating possible
reception. At one site, we aimed for maximum signal strength and
couldn't get 8-VSB pictures until we'd repositioned to a direction that
was about 12 dB down from the maximum reading. A spectrum analyzer may
not be much better. One site appeared to have a relatively flat
spectrum well above threshold, but tightly spaced nulls seem to have
been enough to prevent 8-VSB reception.
- As in real estate, so, too, in DTV reception: location, location,
location! At ground level (receive antennas at roughly head height),
EVERYTHING affected the look of the channel on the spectrum analyzer --
passing trucks, blowing leaves, and us. Coming near the twin-lead when
rotating the directional antenna would sometimes have more effect than
the rotation itself. Sometimes the RF seemed to change for no apparent
reason -- traffic elsewhere? A plane we couldn't see or hear? Who
knows? On the other hand, nine stories up, the channel was pretty
solid. This is all based on looking at the spectrum analyzer, not
pictures. The spectrum analyzer was neither COFDM nor 8-VSB, and both
modulation schemes used the same channel. We normally don't notice
these things because peaky, ever-changing analog spectra don't show them
well. It's pretty easy to see, however, when the supposed-to-be-flat
digital transmission isn't.
- WITH THE RECEIVERS BEING USED, where signal strength was the
issue, there was little to choose between 8-VSB and COFDM. The first
reception site I selected was actually inside the Sinclair transmitter
building, almost directly under the tower. We were at the end of a long
hallway, the only opening being in the opposite direction from the
tower. Neither system could reliably lock up. With the steel door to
the transmitter room opened, enough signal leaked out to make either
system lock in some antenna position but not all positions. I am told
(but did not confirm -- my choice, due to scheduling; the Sinclair
people were willing) that results were similar in fringe areas; the
signal margins there (determined by adding attenuation) were comparable
between 8-VSB and COFDM.
- In more ordinary locations, where the issue was not signal
strength, COFDM beat 8-VSB hands down. In places where we found either
no reception or a very narrow angular range for 8-VSB (50 degrees or
smaller), COFDM was either good through the full 360 degrees or could
handle a very broad swath (at least 135 degrees). This is comparing the
better of the two 8-VSB receivers with the poorer of the two COFDMreceivers.
- In an indoor location in a high-rise, with an east-facing window
and the transmitter to the north, COFDM locked up without difficulty in
any antenna position or location. Moving around and attempting to
shield the antenna with my body could not make reception fail. In the
same location, I could not get 8-VSB reception anywhere. The Sinclair
personnel, who had been there previously, showed me one place in the
room where they'd found where they could make it work. It required
dismounting the directional antenna from the tripod and placing it
against the window on a broad, horizontal aluminum architectural panel;
8-VSB reception was stable in that position. For the heck of it, I
tried getting reception just by touching the antenna input at that
location. I COULD do it, on one COFDM receiver, but not very reliably.
- Please note that DTV reception is currently a slightly tricky
beast even for COFDM. There were low (ground-level) locations and
antenna positions where there appeared to be sufficient signal strength
but where one of the COFDM receivers would not produce reliable pictures
until the antenna position was changed. Of course, there is every
reason to expect that both 8-VSB and COFDM receivers will improve.
- For what it's worth, two other people noticed me carrying a
spectrum analyzer out of an apartment building and asked if we were
testing HDTV! It turns out they were installing a satellite receiver.
They are based in Virginia and said that they had made one HDTV
installation and that it was successful, but it wasn't clear if it
included a DTV decoder.
more..........................
digitaltelevision.com |
